Hanabi

The 'Hanabi' dataset is a benchmark used to evaluate cooperative multi-agent reasoning in the context of the cooperative card game Hanabi, which involves strategic communication and decision-making among players.

Papers using Hanabi (6)

Don't Make the LLM Read the Graph: Make the Graph Think2026Equivariant Networks For Zero-shot Coordination2022 · 3 citesLearned Belief Search: Efficiently Improving Policies In Partially Observable Settings2021 · 2 citesTowards Few-shot Coordination: Revisiting Ad-hoc Teamplay Challenge In The Game Of Hanabi2023Any-Play: An Intrinsic Augmentation for Zero-Shot Coordination2022 · 4 citesOn-the-fly Strategy Adaptation for ad-hoc Agent Coordination2022 · 2 cites